I'm amplifying the output of a 100mohm 4 terminal (kelvin) shunt with the AD8253. I'm using a lab quality calibrator as the signal (current) source. Problem I'm having is when injecting a DC current through the shunt; I have a lot of 50Hz coupling into the output of the AD8253. Measuring with a scope accross the inputs shows no 50Hz content. I have tried a variety of bias resistors from inputs to AGND (as low as 10k). Injecting a 50Hz current results in a output signal that beats with mains.
How can I minimize this effect with the AD8253?